, Volume, Number 4, p.138-14, 23 EXPERIMENTAL STUDIES ON THE EFFECT OF THE FIRE POSITION ON PLUME ENTRAINMENT IN A LARGE SPACE Yuanzhou Li, Ran Huo, Liang Yi and Guodong Wang State Key Laboratory of Fire Science, University of Science and Technology of China, Hefei 2326, Anhui, China W.K. Chow Department of Building Services Engineering, The Hong Kong Polytechnic University, Hong Kong, China ABSTRACT Experiments were carried out in a full-scale burning facility to study the characteristics of plume entrainment for different fires located at the center, near wall and at the corner in a large space. From the results of the smoke layer height, it was found that the plume entrainment rate was the highest for fires at the center, and the lowest for fires at the corner. The results obtained basically agreed with those in the literature, but the entrainment coefficient has to be modified for different positions of the fire plume. Keywords: smoke plume, entrainment, smoke layer height 1. INTRODUCTION Smoke produced from a fire has a relatively strong obscuration, and contains many toxic substances. It is the major hazard to people in building fires. It is very important to study the smoke production and development for better smoke control in buildings. Smoke production rate has a large effect on smoke development, and it depends on the amount of air entrained by the plume. In a fire, the combustion products will move upward due to buoyancy. During this process, the surrounding cool air will be entrained into the smoke and a smoke plume will be formed. The entrained air is the major source for smoke production. Thus, investigating the plume entrainment characteristics would provide a basis for studying smoke control. When the fire source is located at different positions inside a building, the amount of air entrained will be different due to different entrainment conditions; and different forms of plume will be formed. Common plume models include axisymmetric plume (non-restricted plume), wall plume and corner plume. In this paper, based on the empirical formula in the literature, the entrainment characteristics of smoke plume for fires at the center, near wall and at the corner were studied by full-scale burning tests. 2. ENTRAINMENT FOR FIRES AT DIFFERENT POSITIONS Axisymmetric plume entrainment is not affected by walls or other obstructions, so that air can be entrained from any directions along the height. When the fire source is located at the center in a large space, since it is far from the walls, smoke development is not greatly affected by the walls so that air can be entrained freely. Under this condition, the plume can be approximated as an axisymmetric plume, which can be simplified as a cone from a virtual point source [1]. Based on this, plume experiments were conducted by Zukoski on fires with radius.1 m to. m. The following mass entrainment formula was proposed [2]: 1/ 3 / 3 & (1) m =.71 Q c z where m& is the mass flow rate of the entrained air (kgs -1 ), Q c is the heat release rate (kw) due to convection, and z is the height of the plume above floor (m). Wall plume When the fire source is located near the wall, air is mainly entrained on the other side far from the wall. Under this situation, it can be approximated that the smoke entrainment rate is half of that of a nonrestricted plume from two identical fire sources [3,4]. Substituting 2Q (two times the heat release rate Q) into equation (1) and then dividing the equation by two gives the mass entrainment of a wall plume: 1 / 3 / 3 & (2) m =.4 Q c Z Axisymmetric plume is also called non-restricted plume or free plume, meaning that the smoke 138
Corner plume When the fire source is located at the corner, air entrainment is affected by two sidewalls, and so the entrainment rate would be even smaller. Under this condition, the entrainment rate can be approximated as 1/4 of that of an axisymmetric plume. Substituting 4Q into equation (1) and then dividing the equation by 4 gives the mass entrainment of a corner plume: 1 / 3 / 3 & (3) m =.28 Q c Z In the experiments, assuming that the mass entrainment rate can be expressed by: 1/ 3 / 3 & = CQc Z (4) m where C is a constant of which the value is to be determined. Smoke layer height By using the zone concept, simplifying the conservation equations of mass, energy and momentum, the equation for smoke layer height can be obtained []. 1 / 3 2 / 3 2 / 3 2CQ c Z H = t = kt () 3 A ρ By fitting a line on the plot of Z -2/3 t of the smoke filling experiments, the gradient k can be obtained, and then the value of C can be approximated. C = (6) 2 3kA ρ (.7Q ) 1 / 3 3. EXPERIMENTS Experiments were carried out at the full-scale burning facility [,6] jointly built by the University of Science and Technology of China and The Hong Kong Polytechnic University in Hefei, Anhui, China. Fifteen sets of experiments were conducted with the fires at the center, near wall and at the corner in the large space. Five different heat release rates were used with the oil pan size varying from.6 m to 1. m. The experimental conditions are listed in Table 1. All the vents were closed in the experiments, but there were openings at the bottom of the room for air supply. Four openings of size.26 x 1.6 m 2 were located at the bottom of the eastern side, all of them were opened throughout the experiments. There was another opening of size 4 x.3 m 2 at the bottom of the western side, which remained closed during the experiment. Table 1: Experimental conditions Expt. no. Position of oil pan Size of oil pan 1 center.6 m x.6 m 2 center.7 m x.7 m 3 center.8 m x.8 m 4 center.9 m x.9 m center Two.7 m x.7 m 6 near wall.6 m x.6 m 7 near wall.7 m x.7 m 8 near wall.8 m x.8 m 9 near wall.9 m x.9 m 1 near wall Two.7 m x.7 m 11 corner.6 m x.6 m 12 corner.7 m x.7 m 13 corner.8 m x.8 m 14 corner.9 m x.9 m corner Two.7 m x.7 m 4. RESULTS AND DISCUSSION 4.1 Experimental Observations Once the oil was ignited, it burned rapidly into a pool fire and a large amount of hot smoke was produced. When the fire was located at the center, the rising smoke appeared as an axisymmetric plume. About 2 s later, smoke reached the ceiling (the larger the heat release rate, the faster the rate) and formed a ceiling jet. The ceiling jet continued to spread horizontally and upon hitting the sidewall, it turned to spread downwards and formed an antibuoyant wall jet. It then rose again after descending for a certain distance. Along with the smoke development towards the center of the space, a smoke layer with rather uniform thickness was gradually formed. When the fire was located near the wall, smoke appeared as semi-conical. The rising rate was faster than when the fire was at the center, and smoke reached the ceiling in a shorter time. The smoke entrainment rate was reduced significantly. When the fire was located at the corner, smoke appeared as 1/4 of a cone. It had the fastest rising rate, the shortest time to reach the ceiling and also, the smallest smoke entrainment rate. 4.2 Results and Discussion 4.2.1 Comparison of smoke layer height for different fire sizes The results of smoke layer height for different fire sizes with the fire at the center, near wall and at the 139
corner are shown in Fig. 1. From the figures, similar trends of smoke layer height for the three different positions are observed. When the oil pan size increased, the smoke descending rate also increased. The descending rate was faster in the Smoke 烟气层高度 layer height (m) (m) 3 2 1 1 2 3 4 beginning, and then gradually decreased when smoke had descended to 1 m. It can also be seen that the smoke descending rate was the fastest when the fire was located at the center, and the rate was the slowest for the fire located at the corner..6.6.7.7.8.8.9.9 Two 两个.7.7 (a) At the center Smoke 烟气层高度 layer height (m) (m) Smoke 烟气层高度 layer height (m) (m) 3 2 1 3 2 1 时间 Time (s) (s).6.6.7.7.8.8.9.9 两个 Two.7.7 2 4 6 8 Time (s) 时间 (s).6.6.7.7.8.8.9.9 Two 两个.7.7 (b) Near wall (c) At the corner 1 2 3 4 6 7 Time 时间 (s) (s) Fig. 1: Comparison of smoke layer height for fires at the same position but of different sizes 14
4.2.2 Comparison of smoke layer height for fires at different positions The results of the smoke layer height under five different fire sizes at the center, near wall and at the corner are shown in Fig. 2. It can be seen that the smoke descending rate was the faster when the fire was located at the center, and the rate was the slowest for the fire located at the corner. For the case with the oil pan of size.6 m x.6 m, it took about 2 s for the smoke to descend to 7 m for the fire at the center, but it took about 41 s for the case with the fire near wall, and 6 s for the case with the fire at the corner. As the oil pan size increased, the smoke descending rate increased, and the effect of the fire position also decreased. For example, in the case with two oil pans of size.7 m x.7 m, it took about 2 s for the smoke to descend to m for the fire at the center, but it took about 3 s for the case with the fire near wall, and s for the case with the fire at the corner. The calculation results by substituting Zukoski s plume model into the equation for smoke layer height are shown as the curved line in the figure. It can be seen that the results agreed fairly well. The calculated smoke descending rates are generally faster than the experimental results. Smoke layer height (m) 烟气层高度 (m) 3 2 1 中央 Center 墙边 Near wall 墙角 Corner 中央 Center 墙边 Near wall 墙角 Corner 1 2 3 4 6 7 Time 时间 (s) (a).6 m.6 m oil pan Smoke layer height (m) 烟气层高度 (m) 3 2 1 中央 Center 墙边 Near wall 墙角 Corner 中央 Center 墙边 Near wall 墙角 Corner 1 2 3 4 6 7 Time 时间 (s) (s) (b).7m.7m oil pan Fig. 2: Comparison of smoke layer height for fires of the same size but at different positions 141
Smoke 烟气层高度 layer height (m) (m) 3 2 1 Smoke 烟气层高度 layer (m) height (m) Smoke 烟气层高度 layer height (m) (m) 3 2 1 3 2 1 1 2 3 4 6 7 Time 时间 (s) (s) (c).8 m.8 m oil pan 1 2 3 4 6 7 Time 时间 (s) (s) (d).9 m.9 m oil pan 1 2 3 4 6 7 Time (s) 时间 (s) (e) Two.7 m.7 m oil pans Fig. 2 (cont d): Comparison of smoke layer height for fires of the same size but at different positions 142
4.3 Improvement of C Value The values of Z -2/3 t for fires at the center, near wall and at the corner are plotted in Figs. 3 to. According to equation (6), the value of the entrainment coefficient C for each experiment can be calculated from the gradient k, heat release rate Q and the plane area A. In this study, the area of the large space is 266.6 m 2, the values of heat release rate Q, gradient k and entrainment coefficient C for each experiment are listed in Table 2. The average plume entrainment coefficients for fires at different positions are: At the center: C a =.77 +.77 +.79 +.723 +.732 =.716 Near wall: C w =.436 +.41+.46 +.4 +.433 =.42 At the corner: C c =. +.27 +. +.236 +. =..7.6..4.3.6*.6.7*.7.8*.8.9*.9 Two 两个.7*.7.2.1 1 2 3 4 Time 时间 (s) Fig. 3: Processed experimental data for fires at the center..4.3.6.6.7.7.8.8.9.9 两个 Two.7.7.2.1. 1 2 3 4 6 7 Time 时间 (s) Fig. 4: Processed experimental data for fires near wall 143
.3..2...6.7.7.8.8.9.9 两个 Two.7.7.1.. 1 2 3 4 6 7 时间 Time (s) (s) Fig. : Processed experimental data for fires at the corner Table 2: Values of heat release rate Q, gradient k and entrainment coefficient C in each experiment Expt. no. Heat release rate (kw) Gradient k 1 336.9.77 2 612..11.77 3 69.11.79 4 94.14.723 1142.4.14.732 6 298.2.4.436 7 9.6.6.41 8.6.46 9 783.7.4 1 124.8.433 11 298.2.3. 12 4.4.27 13 676.2.4. 14 793.8.4.236 132... CONCLUSION Fifteen sets of experiments were carried out with five different heat release rates and three different fire positions in a large space. It was found that the smoke descending rate is highly related to the heat C release rate and the position of the fire source. The higher the heat release rate, the faster the smoke descending rate, which is approximately proportional to 1/3 power of the heat release rate. Also, the plume entrainment rates are different for different fire positions. The plume entrainment rates at the center, near wall and at the corner are.72,.42, and. respectively. The entrainment coefficients for fires near wall and at the corner are slightly smaller than the values obtained from mirror-reflection analysis. ACKNOWLEDGEMENT The work is supported by National Key Basic Research Special Funds of China under Grant No. 21CB4964 and the President of The Hong Kong Polytechnic University with account number A-78. REFERENCES 1. B.R. Morton, G. Taylor and J.S. Turner, Turbulent gravitational convection from maintained and instantaneous sources, Proceedings of the Royal Society of London, A234, pp. 1-23 (196). 2. E. Zukoski, Entrainment in fire plumes, Fire Safety Journal, Vol. 3, No. 2, pp. 17-114 (1983). 3. R.L. Alpert and E.J. Ward, Evaluation of unsprinklered fire hazard, Fire Safety Journal, Vol. 1, pp. 127-143 (1984). 4. F.W. Mowerer and B. Williamson, Estimating room temperature from fires along walls an in 144
corners, Fire Technology, Vol. 23, No. 2, pp. 133-14 (1987).. R. Huo, Y.Z. Li, X.F. Jin and W.C. Fan, Smoke filling in a large space, Combustion Science and Technology, Vol. 7, No. 3, pp. 219-222 (21). 6. W.K. Chow, Y.Z. Li, E. Cui and R. Huo, Natural smoke filling in atrium with liquid pool fires up to 1.6 MW, Building and Environment, Vol. 36, No. 1, pp. 121-127 (21). 14